In this notebook we aim to answer the question:

What products can we strategize to different customer segments to improve their engagement?

1. Analyse commodities that enjoyed by all segments
   • Department Engagements and Revenues
   • Top 5 commodities in all departments
2. Analyse commodities that resonates with family households and age band
   • Baby Care
   • Dairy Products
3. Analyse commodities that resonates with single female and single male
   • Department Engagements
   • Top 5 Grocery and Produce commodities

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import plotly.graph_objects as go
import plotly.io as pio 
import seaborn as sns

df = pd.read_csv("dataset_2019_2022_new.csv")

df.transaction_date = pd.to_datetime(df.transaction_date)
df['year'] = df.transaction_date.dt.year 

Analyze commodities that is enjoyed by all customer segments

Department Revenues and Engagement

df2 = df.groupby(['department']) \
    .agg(
        engagement=('commodity', 'count'),
        revenue=('price',sum)
    ).reset_index()
df2

	department	engagement	revenue
0	Cosmetics	455	2528.65
1	Deli	3635	15085.84
2	Floral	345	2588.94
3	Grocery	25515	63534.81
4	Meat	10855	52392.34
5	Nutrition	2471	8809.72
6	Pastry	4636	13725.72
7	Pharmaceutical	11638	44603.66
8	Produce	16163	37207.43
9	Salad Bar	484	1580.44
10	Seafood	1494	8922.92

fig = go.Figure()

fig.add_trace(go.Scatter(x=df2['engagement'], y=df2['revenue'], mode='markers',
                        marker=dict(color=df2['revenue']), text=df2['department'])) # hover text goes here


highest = df2.loc[df2['revenue'].idxmax(), 'department']
lowest = df2.loc[df2['revenue'].idxmin(), 'department']

fig.add_annotation(x=df2.loc[df2['engagement'].idxmax(), 'engagement'], 
                   y=df2.loc[df2['revenue'].idxmax(), 'revenue'], 
                   text=highest, showarrow=True, arrowhead=1)
fig.add_annotation(x=df2.loc[df2['engagement'].idxmin(), 'engagement'], 
                   y=df2.loc[df2['revenue'].idxmin(), 'revenue'], 
                   text=lowest, showarrow=True, arrowhead=1)

fig.update_layout(
    title='Department - Engagement vs Revenue',
    yaxis=dict(title='Revenue'),
    xaxis=dict(title='Engagement')
)

fig.show()

This scatter plot illustrates the relationship between engagement and revenue for different departments. Grocery stands out with the highest engagement and highest revenue. Other stands out are Meat, Pharmaceutical, and Produce.

Based on the insight provided, we can allocate marketing efforts and resources to categories with high engagement and revenue to further boost sales and customer satisfaction.

Top 5 commodities on each department

#top 5 commodities by department
df3 = df.groupby(['department', 'commodity']) \
    .agg(
        engagement=('commodity', 'count'),
        revenue=('price',sum)
    ).reset_index()

top5_df = pd.concat(
    [df3[df3.department == d] \
         .sort_values('revenue', ascending=False) \
     .head(5) for d in df3.department.unique()]).reset_index(drop=True)
top5_df

	department	commodity	engagement	revenue
0	Cosmetics	Makeup and treatment	367	1360.05
1	Cosmetics	Coupon/misc items	20	799.80
2	Cosmetics	Cosmetic accessories	12	155.88
3	Cosmetics	Bath	48	152.96
4	Cosmetics	Fragrances	8	59.96
5	Deli	Deli meats	1530	5912.61
6	Deli	Cheese	795	3312.77
7	Deli	Chicken/poultry	388	1883.68
8	Deli	Prepared food	260	1279.49
9	Deli	Salad	336	1161.36
10	Floral	Floral plants	288	2236.51
11	Floral	Roses	33	310.67
12	Floral	Floral accessories	24	41.76
13	Grocery	Soft drinks	1724	4194.49
14	Grocery	Fluid milk products	1378	3347.96
15	Grocery	Cheese	1267	3056.22
16	Grocery	Frozen meat	1028	2879.74
17	Grocery	Baked bread/buns/rolls	1259	2370.49
18	Meat	Beef	2940	17297.40
19	Meat	Lunch meat	2075	5572.08
20	Meat	Pork	684	5027.56
21	Meat	Chicken	796	4283.37
22	Meat	Meat - Other	536	3426.20
23	Nutrition	Refrigerated	804	3152.84
24	Nutrition	Fitness/diet	236	977.56
25	Nutrition	Frozen	220	901.04
26	Nutrition	Cereal/breakfast	136	636.00
27	Nutrition	Water	275	473.36
28	Pastry	Cakes	626	3461.57
29	Pastry	Bread	1240	2810.56
30	Pastry	Breakfast sweets	1317	2732.00
31	Pastry	Cookies	408	1702.64
32	Pastry	Pies	328	1435.84
33	Pharmaceutical	Cigarettes	569	4826.62
34	Pharmaceutical	Candy	2862	4603.99
35	Pharmaceutical	Hair care products	606	2104.35
36	Pharmaceutical	Oral hygiene products	576	1944.93
37	Pharmaceutical	Diapers	177	1853.37
38	Produce	Salad	1125	2999.69
39	Produce	Potatoes	947	2875.05
40	Produce	Apples	942	2753.67
41	Produce	Berries	734	2675.50
42	Produce	Tropical fruit	2145	2570.62
43	Salad Bar	Salad	480	1549.04
44	Salad Bar	Produce supplies	4	31.40
45	Seafood	Seafood-frozen	872	5762.80
46	Seafood	Seafood-fresh	432	2621.68
47	Seafood	Seafood-misc	150	314.68
48	Seafood	Frozen boxed/grocery	40	223.76

top5_df.groupby('department').agg(
    total_rev=('revenue',sum),
    total_eng=('engagement',sum)
).reset_index()

	department	total_rev	total_eng
0	Cosmetics	2528.65	455
1	Deli	13549.91	3309
2	Floral	2588.94	345
3	Grocery	15848.90	6656
4	Meat	35606.61	7031
5	Nutrition	6140.80	1671
6	Pastry	12142.61	3919
7	Pharmaceutical	15333.26	4790
8	Produce	13874.53	5893
9	Salad Bar	1580.44	484
10	Seafood	8922.92	1494

tmp = top5_df.groupby('department').agg(
    total_rev=('revenue',sum),
    total_eng=('engagement',sum)
).reset_index()

plt.figure(figsize=(10, 5))
sns.set(style='whitegrid')
ax = sns.barplot(x='department', y='total_rev', data=tmp, palette='Set3')
plt.xticks(rotation=45)
plt.show()

tmp = top5_df.groupby('department').agg(
    total_rev=('revenue',sum),
    total_eng=('engagement',sum)
).reset_index()

plt.figure(figsize=(10, 5))
sns.set(style='whitegrid')
ax = sns.barplot(x='department', y='total_eng', data=tmp, palette='Set3')
plt.xticks(rotation=45)
plt.show()

#revenue bar chart

fig = go.Figure()

for d in top5_df.commodity.unique():
    tmp = top5_df[top5_df.commodity==d].copy()
    fig.add_trace(go.Bar(name=d,x=tmp['department'],y=tmp['revenue'],
                        hovertext=tmp['revenue'].round(2)))
    
fig.update_layout(barmode='stack')
fig.update_layout(title ='Top 5 Commodities on each Department - Revenue',
        yaxis=dict(
            title='Revenue'
        ))
fig.show()

#engagement bar chart

fig = go.Figure()

for d in top5_df.commodity.unique():
    tmp = top5_df[top5_df.commodity==d].copy()
    fig.add_trace(go.Bar(name=d,x=tmp['department'],y=tmp['engagement'],
                        hovertext=tmp['engagement'].round(2)))
    
fig.update_layout(barmode='stack')
fig.update_layout(title ='Top 5 Commodities on each Department - Engagement',
        yaxis=dict(
            title='Engagement'
        ))
fig.show()

The analysis of the top 5 commodities on each department based on both revenue and engagement reveals valuable insights into the departmental performance, emphasizing strategic opportunities between revenue generation and customer engagement.

1. Revenue Focused Analysis
   • The top 5 commodities on the Meat department, collectively contribute significanlty to the department's revenue, exceeding $35,000. This underscores strong consumer demand and succesful sales within the Meat Department.
   • While the top 5 commodities contribute notable revenue in these departments, they fall short of the Meat Department's performance, with revenue totaling over $15,000 each.
2. Engagement-Focused Analysis:
   • Similarly, the Meat Department secures the highest customer engagement, amassing over 7,000 engagements. This reflects the department's ability to effectively capture customer interest and interactions.
   • Following closely, the Grocery Department records approximately 6,500 engagements, showcasing its successful consumer appeal.

Analyze commodities that resonates with family households and age band

Baby Care

# Average engagement frequency from baby items per household
baby_df = df[df.commodity.str.lower().str.contains('baby|diapers')] \
    .groupby(['year','household_type','commodity']) \
    .agg(engagement=('commodity','count')).reset_index()

baby_df.groupby(['household_type','commodity']).agg(avg_engagement=('engagement','mean')).reset_index()

	household_type	commodity	avg_engagement
0	1 adult with kids	Baby foods	18.500000
1	1 adult with kids	Baby hbc	15.000000
2	1 adult with kids	Diapers	10.500000
3	2 adults with kids	Baby foods	27.250000
4	2 adults with kids	Baby hbc	21.500000
5	2 adults with kids	Diapers	13.500000
6	2 adults with no kids	Baby foods	17.250000
7	2 adults with no kids	Baby hbc	16.000000
8	2 adults with no kids	Diapers	10.500000
9	Single female	Baby foods	11.666667
10	Single female	Baby hbc	7.000000
11	Single female	Diapers	2.500000
12	Single male	Baby foods	17.750000
13	Single male	Baby hbc	13.250000
14	Single male	Diapers	7.250000

data_baby = []
baby_color = {
    'Baby foods': '#ff9999',
    'Baby hbc': '#cc99ff',
    'Diapers': '#99ffff'
}

for d in baby_df.commodity.unique():
    tmp = baby_df[baby_df.commodity==d].groupby(['household_type','commodity']) \
        .agg(avg_engagement=('engagement','mean')).reset_index()
    tmp['avg_engagement'] = tmp['avg_engagement'].round(0)
    data_baby.append(go.Bar(x=tmp.household_type, 
                            y=tmp.avg_engagement, 
                            name=d, 
                            text=tmp.avg_engagement, 
                            marker_color=baby_color[d]))
    
fig = go.Figure(
    data=data_baby,
    layout = go.Layout(
        title ='Baby Items Engagement Frequency by Household Type',
        yaxis=dict(
            title='Average Engagement'
        )
    )
)


fig.show()

Household Type: The grouped bar chart reveals intriguing patterns in average engagement accross baby items per household.

Insights:

1. Among households with kids, both single and couples, Baby Foods exhibit relatively high average engagement. This suggests that families with children, regardless of household composition, show a strong interest in this commodity.

2. Similar to Baby Foods, Baby HBC products also demonstrate substantial engagement across different household types. This implies that products catering to the health and beauty needs of infants and children hold a consistent appeal.

3. Diapers, a fundamental necessity for families with young children, receive significant engagement from households with kids. Interestingly, households with two adults and kids show a higher engagement with this commodity compared to households with just one adult and kids.

4. Both Single Females and Single Males exhibit lower engagement across all commodities compared to households with kids. Baby Foods garner relatively higher engagement among Single Males, potentially indicating their role as caregivers. Diapers, understandably, have the least engagement across both Single Female and Single Male households.

5. Households without Kids: Households with two adults and no kids display moderate engagement with Baby Foods and Baby HBC, suggesting that these products might cater to specific needs even in households without children. Diapers exhibit lower engagement in these households, reflecting their lack of relevance.

# Average engagement frequency from baby items per age band
baby_df2 = df[df.commodity.str.lower().str.contains('baby|diapers')] \
    .groupby(['year','age_band','commodity']) \
    .agg(engagement=('commodity','count')).reset_index()
baby_df2.groupby(['age_band','commodity']).agg(avg_engagement=('engagement','mean')).reset_index()

	age_band	commodity	avg_engagement
0	19-24	Baby foods	45.75
1	19-24	Baby hbc	37.50
2	19-24	Diapers	21.50
3	25-34	Baby foods	15.00
4	25-34	Baby hbc	12.25
5	25-34	Diapers	8.50
6	35-44	Baby foods	4.75
7	35-44	Baby hbc	6.00
8	35-44	Diapers	5.50
9	45-54	Baby foods	24.00
10	45-54	Baby hbc	18.50
11	45-54	Diapers	11.50

data_baby = []

for d in baby_df2.commodity.unique():
    tmp = baby_df2[baby_df2.commodity==d].groupby(['age_band','commodity']) \
        .agg(avg_engagement=('engagement','mean')).reset_index()
    tmp['avg_engagement'] = tmp['avg_engagement'].round(0)
    data_baby.append(go.Bar(x=tmp.age_band, 
                            y=tmp.avg_engagement, 
                            name=d, text=tmp.avg_engagement,
                            marker_color=baby_color[d]))
    
fig = go.Figure(
    data=data_baby,
    layout = go.Layout(
        title ='Baby Items Engagement Frequency by Age Band',
        yaxis=dict(
            title='Average Engagement'
        )
    )
)

fig.show()

Age Band: These insights underscore the need for SuperFoodsMax to tailor its marketing strategies and product offerings according to the distinct preferences and priorities of different age bands. For the 19-24 age group, emphasizing baby care commodities and providing resources for young parents could enhance engagement. In the 25-34 age band, SuperFoodsMax can explore strategies that cater to potential parenting considerations while still valuing personal well-being. For older age bands, focusing on broader wellness offerings while acknowledging reduced engagement with baby care products could be a strategic approach.

Dairy Products

dairy_df=df[df.commodity.str.lower().str.contains('milk|cheese|butter|egg|yogurt|cream')] \
    .groupby(['year','household_type','commodity','brand']) \
    .agg(
    engagement=('commodity','count'),
    revenue=('price',sum)).reset_index()

dairy_df

	year	household_type	commodity	brand	engagement	revenue
0	2019	1 adult with kids	Butter	national	2	10.64
1	2019	1 adult with kids	Butter	private	4	9.95
2	2019	1 adult with kids	Canned milk	national	1	2.00
3	2019	1 adult with kids	Cheese	national	88	331.66
4	2019	1 adult with kids	Cheese	private	42	92.96
...	...	...	...	...	...	...
306	2022	Single male	Ice cream/milk/sherbets	private	4	8.65
307	2022	Single male	Milk by-products	private	5	6.15
308	2022	Single male	Peanut butter/jelly/jams	private	2	3.99
309	2022	Single male	Yogurt	national	11	15.24
310	2022	Single male	Yogurt	private	8	4.73

311 rows × 6 columns

dairy_df.groupby(['household_type','commodity']).agg(avg_engagement=('engagement','mean')).reset_index()

	household_type	commodity	avg_engagement
0	1 adult with kids	Butter	3.750000
1	1 adult with kids	Canned milk	1.000000
2	1 adult with kids	Cheese	63.875000
3	1 adult with kids	Eggs	17.000000
4	1 adult with kids	Fluid milk products	42.375000
5	1 adult with kids	Ice cream/milk/sherbets	10.375000
6	1 adult with kids	Milk by-products	11.125000
7	1 adult with kids	Peanut butter/jelly/jams	6.125000
8	1 adult with kids	Yogurt	24.250000
9	2 adults with kids	Butter	5.375000
10	2 adults with kids	Canned milk	2.000000
11	2 adults with kids	Cheese	75.625000
12	2 adults with kids	Eggs	17.857143
13	2 adults with kids	Fluid milk products	51.250000
14	2 adults with kids	Ice cream/milk/sherbets	14.375000
15	2 adults with kids	Milk by-products	15.500000
16	2 adults with kids	Peanut butter/jelly/jams	9.285714
17	2 adults with kids	Yogurt	28.125000
18	2 adults with no kids	Butter	4.875000
19	2 adults with no kids	Canned milk	1.250000
20	2 adults with no kids	Cheese	53.500000
21	2 adults with no kids	Eggs	11.000000
22	2 adults with no kids	Fluid milk products	40.750000
23	2 adults with no kids	Ice cream/milk/sherbets	8.875000
24	2 adults with no kids	Milk by-products	10.000000
25	2 adults with no kids	Peanut butter/jelly/jams	7.000000
26	2 adults with no kids	Yogurt	24.125000
27	Single female	Butter	2.750000
28	Single female	Canned milk	1.000000
29	Single female	Cheese	24.750000
30	Single female	Eggs	5.500000
31	Single female	Fluid milk products	18.333333
32	Single female	Ice cream/milk/sherbets	4.875000
33	Single female	Milk by-products	5.000000
34	Single female	Peanut butter/jelly/jams	2.666667
35	Single female	Yogurt	8.750000
36	Single male	Butter	4.000000
37	Single male	Canned milk	1.000000
38	Single male	Cheese	40.000000
39	Single male	Eggs	10.428571
40	Single male	Fluid milk products	24.125000
41	Single male	Ice cream/milk/sherbets	7.375000
42	Single male	Milk by-products	8.142857
43	Single male	Peanut butter/jelly/jams	4.142857
44	Single male	Yogurt	15.625000

data_dairy = []

for d in dairy_df.commodity.unique():
    tmp = dairy_df[dairy_df.commodity==d].groupby(['household_type','commodity']) \
        .agg(avg_engagement=('engagement','mean')).reset_index()
    tmp['avg_engagement'] = tmp['avg_engagement'].round(0)
    data_dairy.append(go.Bar(x=tmp.household_type, y=tmp.avg_engagement, name=d, text=tmp.avg_engagement))
    
fig = go.Figure(
    data=data_dairy,
    layout = go.Layout(
        title ='Dairy Products and Eggs Engagement Frequency by Household Type',
        yaxis=dict(
            title='Average Engagement'
        )
    )
)


fig.show()

Insight:

1. There is a huge Cheese engagement in all household types. Cheese is prominently favored by all household types.
2. Fluid milk products is also the second favored among all household types.
3. Couples (with or without kids) seem to favor butter in their purchases. Single female and Single male household shows comparatively lower engagement with butter.
4. Yogurt seems to resonate more with households comprising couples, especially those with children.

Recommendation:

1. We can target promotion of cheese among 'Single females', or yogurt for household with kids.

dairy_df2 = dairy_df.groupby(['commodity','brand']).agg(total_rev=('revenue',sum)) \
    .sort_values(['commodity','total_rev'], ascending=False).reset_index()

dairy_df2

	commodity	brand	total_rev
0	Yogurt	national	877.61
1	Yogurt	private	184.30
2	Peanut butter/jelly/jams	national	307.50
3	Peanut butter/jelly/jams	private	276.85
4	Milk by-products	private	464.18
5	Milk by-products	national	176.75
6	Ice cream/milk/sherbets	national	668.64
7	Ice cream/milk/sherbets	private	405.31
8	Fluid milk products	private	3100.48
9	Fluid milk products	national	247.48
10	Eggs	private	512.47
11	Eggs	national	52.67
12	Cheese	national	4049.76
13	Cheese	private	2319.23
14	Canned milk	national	24.27
15	Canned milk	private	8.14
16	Butter	private	304.74
17	Butter	national	153.76

dairy_df2.pivot(index='commodity', columns='brand', values='total_rev') \
    .plot(kind='bar', stacked=True, title='Dairy Commodities Revenue by Brands', figsize=(12,6))


plt.xlabel(None)
plt.ylabel('Revenue')
plt.draw()
plt.savefig('images\dairy_by_brand.png',dpi=100)

plt.show()

Insight:

1. Preference for national brands for the following items: Cheese, Yogurt, Peanut Butter/Jelly/Jams, and Ice cream/Milk/Sherbets products.
2. Private brands are leading in Butter, Fluid Milk Products, and Milk by-products.

Recommendation:

1. To boost revenue in categories like Canned Milk, and butter SuperFoodsMax can implement promotions to encourage purchases.
2. While national brands are popular, SuperFoodsMax must maintain a balance between private brands and national brands across various categories can provide customers with more choices and appeal to different segments of the market.

Analyze commodities that resonates with single customers

Single Female

sf_df=df[df.household_type=='Single female'].copy()

sf_df.groupby('age_band').agg(count=('customer_id',pd.Series.nunique)).reset_index()

	age_band	count
0	19-24	221
1	25-34	54
2	35-44	2
3	45-54	1
4	55-64	1

tmp = sf_df.groupby('age_band').agg(count=('customer_id',pd.Series.nunique)).reset_index()

fig = go.Figure(
    data=[go.Pie(labels=tmp['age_band'], values=tmp['count'])]
)

fig.show()

sf_df.groupby('department').agg(engagement=('department','count')).reset_index()

	department	engagement
0	Cosmetics	44
1	Deli	303
2	Floral	28
3	Grocery	2268
4	Meat	994
5	Nutrition	217
6	Pastry	402
7	Pharmaceutical	1045
8	Produce	1347
9	Salad Bar	36
10	Seafood	200

tmp = sf_df.groupby('department').agg(engagement=('department','count')).reset_index()
plt.figure(figsize=(10, 5))
plt.title('Department Engagement for Single Female')
sns.set_style("whitegrid")
ax = sns.barplot(x='department', y='engagement', data=tmp, palette='Set2')
plt.xticks(rotation=45)
plt.savefig('images\single_female_dept.png')
plt.show()

dept = ['Grocery','Produce']

#top 5 grocery items and produce of single female

tmp = pd.concat([sf_df[sf_df.department == d] \
            .groupby(['household_type','department','commodity']) \
            .agg(engagement=('commodity','count')) \
            .sort_values(['engagement'], ascending=False) for d in dept]).reset_index()

sf_top5 = pd.concat(
    [tmp[tmp.department == d] \
         .sort_values('engagement', ascending=False) \
     .head(5) for d in tmp.department.unique()]).reset_index(drop=True)
sf_top5

	household_type	department	commodity	engagement
0	Single female	Grocery	Soft drinks	144
1	Single female	Grocery	Baked bread/buns/rolls	134
2	Single female	Grocery	Cheese	121
3	Single female	Grocery	Fluid milk products	110
4	Single female	Grocery	Frozen meat	91
5	Single female	Produce	Tropical fruit	158
6	Single female	Produce	Salad	112
7	Single female	Produce	Apples	91
8	Single female	Produce	Onions	86
9	Single female	Produce	Potatoes	70

plt.figure(figsize=(10, 5))
plt.title('Department Commodity Engagement - Single Female')

sns.set_style("whitegrid")
ax = sns.barplot(x='department', y='engagement', hue='commodity', data=sf_top5, palette='Set3')
ax.set_xticklabels(ax.get_xticklabels(), ha='center')


plt.legend(loc='center left', bbox_to_anchor=(1.0, 0.5))
plt.savefig('images\single_female_dept_commodity.png')
plt.show()

Single Male

sm_df=df[df.household_type=='Single male'].copy()

sm_df.groupby('age_band').agg(count=('customer_id',pd.Series.nunique)).reset_index()

	age_band	count
0	19-24	384
1	25-34	106
2	35-44	8

tmp = sm_df.groupby('age_band').agg(count=('customer_id',pd.Series.nunique)).reset_index()

fig = go.Figure(
    data=[go.Pie(labels=tmp['age_band'], values=tmp['count'])]
)

fig.show()

sm_df.groupby('department').agg(engagement=('department','count')).reset_index()

	department	engagement
0	Cosmetics	85
1	Deli	579
2	Floral	49
3	Grocery	3928
4	Meat	1656
5	Nutrition	339
6	Pastry	699
7	Pharmaceutical	1805
8	Produce	2465
9	Salad Bar	60
10	Seafood	373

tmp = sm_df.groupby('department').agg(engagement=('department','count')).reset_index()
plt.figure(figsize=(10, 5))
plt.title('Department Engagement for Single Male')
sns.set_style("whitegrid")
ax = sns.barplot(x='department', y='engagement', data=tmp, palette='Set2')
plt.xticks(rotation=45)
plt.savefig('images\single_male_dept.png')
plt.show()

#top 5 grocery items and produce of single male

tmp = pd.concat([sm_df[sm_df.department == d] \
            .groupby(['household_type','department','commodity']) \
            .agg(engagement=('commodity','count')) \
            .sort_values(['engagement'], ascending=False) for d in dept]).reset_index()

sm_top5 = pd.concat(
    [tmp[tmp.department == d] \
         .sort_values('engagement', ascending=False) \
     .head(5) for d in tmp.department.unique()]).reset_index(drop=True)
sm_top5

	household_type	department	commodity	engagement
0	Single male	Grocery	Soft drinks	255
1	Single male	Grocery	Cheese	205
2	Single male	Grocery	Baked bread/buns/rolls	194
3	Single male	Grocery	Fluid milk products	193
4	Single male	Grocery	Frozen meat	177
5	Single male	Produce	Tropical fruit	306
6	Single male	Produce	Salad	184
7	Single male	Produce	Vegetables - all others	177
8	Single male	Produce	Onions	162
9	Single male	Produce	Potatoes	155

plt.figure(figsize=(10, 5))
plt.title('Department Commodity Engagement - Single Male')

sns.set_style("whitegrid")
ax = sns.barplot(x='department', y='engagement', hue='commodity', data=sm_top5, palette='Set3')
ax.set_xticklabels(ax.get_xticklabels(), ha='center')


plt.legend(loc='center left', bbox_to_anchor=(1.0, 0.5))
plt.savefig('images\single_male_dept_commodity.png')
plt.show()